The specific aims of this proposal are: 1. To conduct the first extensive demographic and economic study of gay and lesbian households in the United States based on large samples drawn from known sampling frames. 2. To explore a set of research topics that will inform public policy relevant to gays and lesbians. 3. To explore a set of research topics relevant to social science and to describe the similarities and differences among the gay, lesbian, and heterosexual population in choice and outcomes relates to these theories. These issues include assortative mating, family decision making, residential , human capital development, fertility and gender roles. 4. To compare findings from the GSS, NHSLS, and the Census to findings from non-systemic samples of gays and lesbians, in order to assess the properties of these non-systemic samples and to compare inferences one can make using these alternative data sources. Specifically, we seek a clearer understanding of results from American Couples (AC) survey by comparing characteristics of respondents in this survey to those in our standard social science data sets. 5. To develop and exploit the confidential 1-in-6 long form data from the 1990 Census and the matched NSCG in order to achieve the first four aims, and in doing so, to establish the usefulness of confidential Census data products for economic and demographic research more broadly.